Recoverable copper and gold estimates for resource models and ore control routing decisions at Batu Hijau were initially based on flotation testwork from several commercial laboratories. Due to the greater number of copper and gold assay grade available, regression relationships were developed between assay grades and flotation recoveries. Other recovery controlling variables such as oxidation and lithology were included to improve the regression models. With additional drilling, more flotation work was carried out and the regression models were revised to include the new flotation data, also alteration classification and in situ pH._x000D_
The regression models were replaced in favour of spatially derived models for ore control and short range mine planning. To harness local variability in the block model, blast hole flotation testing is carried out in unison with routine grade assaying. Recovery data are processed (kriged) simultaneously with grade assay data prior to ore polygon delineation. Short range mine planning utilises this model in conjunction with a spatial recovery model derived from drill core flotation data._x000D_
Long range mine planning utilises the hybrid' recovery model whereby the globally robust regression recovery models are overprinted by a spatial (kriged) recovery estimate where flotation data density is adequate (Phase II reserve).